Racial, ethnic, and socioeconomic disparities in both health care and health are well documented, but our understanding of the causes of such disparities, and potential solutions, remains incomplete. Previous research has stressed the importance of geography in understanding health care disparities, but these broad regional findings have precluded hospital-level information (or "report cards") on racial or ethnic disparities. Nor do we understand the implications for disparities if Black and white patients from the same neighborhood seek care at different types of hospitals. Our recently completed patient survey suggests large differences by race and ethnicity in beliefs about health care intensity. But we don't know how perceptions of white, Black, Hispanic, and lower income patients translate into the type of care they actually receive. Finally, it is well understood that surgical rates are lower among Black AMI patients, but it is not well understood what the implications are of these lower surgical rates for disparities in health outcomes;and more importantly, what would be the benefits of raising Black surgical rates following AMI. The specific aims are: 1. Develop hospital-level disparity "report cards" for hospital performance with regard to racial, ethnic, or income-based differences in treatment patterns such as quality of care measures and knee and hip replacements using patient-doctor-hospital networks estimated in Core B and Project 1, and quantify the differences in the quality of health care institutions and providers utilized by Black and white patients. 2. Characterize the sorting of patients and providers by race, ethnicity, and socioeconomic status. What is the impact on health outcomes when minority or low-income patients seek care (or are only able to receive care) at largely Black/Hispanic or low-income hospitals, whether through choice or because of hospital closures? 3. Test whether the preferences of patients, as elicited by a recent nation-wide survey of Medicare enrollees, are reflected in the type of health care they subsequently receive, and whether this concordance varies systematically by race, ethnicity, or socioeconomic status using linked Medicare claims data. 4. Estimate the impact of differential racial, ethnic, and income-based cardiac surgical rates on disparities in health outcomes using a large sample (3 million observation) of elderly heart attack patients. Do Black surgical candidates experience better, at the margin, than white surgical candidates?